Perception of dynamic Glass patterns

Abstract

In the mammalian brain, form and motion are processed through two distinct pathways at early stages of visual processing. However, recent evidence suggests that these two pathways may interact. Here we used dynamic Glass patterns, which have been previously shown to create the perception of coherent motion in humans, despite containing no motion coherence. Glass patterns are static stimuli that consist of randomly positioned dot pairs that are integrated spatially to create the perception of a global form, whereas dynamic Glass patterns consist of several independently generated static Glass patterns presented sequentially. In the current study, we measured the detection threshold of five types of dynamic Glass patterns and compared the rank order of the detection thresholds with those found for static Glass patterns and real motion patterns (using random dot stimuli). With both the static Glass patterns and dynamic Glass patterns, detection thresholds were lowest for concentric and radial patterns and highest for horizontal patterns. We also found that vertical patterns were better detected than horizontal patterns, consistent with prior evidence of a “horizontal effect” in the perception of natural scene images. With real motion, detection thresholds were equivalent across all patterns, with the exception of higher thresholds for spiral patterns. Our results suggest that dynamic Glass patterns are processed primarily as form prior to input into the motion system.